math benchmark
FinQA
A large-scale dataset for numerical reasoning over financial data with question-answering pairs written by financial experts, featuring complex numerical reasoning and understanding of heterogeneous representations with annotated gold reasoning programs for full explainability
Updated Aug 7, 2026
Published models3
Registry coverage3
MetricScore
EvidenceB
FinQA leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AM Nova ProAmazon | 77.2% | 100.0% | 3 | C | |
| 2 | AMNova LiteAmazon | 73.6% | 50.0% | 3 | C | |
| 3 | AMNova MicroAmazon | 65.2% | 0.0% | 3 | C |
Score distribution
Top published rows on the benchmark's original scale.
FinQA highlights
The top published results on this benchmark's own scale.
What is FinQA?
Definition and scoring fields from the benchmark registry.
A large-scale dataset for numerical reasoning over financial data with question-answering pairs written by financial experts, featuring complex numerical reasoning and understanding of heterogeneous representations with annotated gold reasoning programs for full explainability
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- FinQA
- Modality
- text
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- finqa|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about FinQA.
Which model scores highest on FinQA?
Nova Pro is currently ranked first with 77.2%.
What does FinQA measure?
A large-scale dataset for numerical reasoning over financial data with question-answering pairs written by financial experts, featuring complex numerical reasoning and understanding of heterogeneous representations with annotated gold reasoning programs for full explainability
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
3 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.